4. Box plot

4.1. 概要

4.2. Plotlyによる作図方法

4.3. MADB Labを用いた作図例

4.3.1. 下準備

import pandas as pd
import plotly.express as px

import warnings
warnings.filterwarnings('ignore')
# 前処理の結果,以下に分析対象ファイルが格納されていることを想定
PATH_DATA = '../../data/preprocess/out/magazines.csv'
# Jupyter Book用のPlotlyのrenderer
RENDERER = 'plotly_mimetype+notebook'
def show_fig(fig):
    """Jupyter Bookでも表示可能なようRendererを指定"""
    fig.show(renderer=RENDERER)
df = pd.read_csv(PATH_DATA)

4.3.2. 雑誌別・作品別の合計連載週数

df_plot = \
    df.value_counts(['mcname', 'cname']).reset_index(name='weeks')
fig = px.box(df_plot, x='mcname', y='weeks', title='雑誌別・作品別の合計連載週数')
show_fig(fig)

Histogramと同様に,2週以上掲載された連載作品に絞って作図してみます.

df_plot = df_plot[df_plot['weeks']>=2].reset_index(drop=True)
fig = px.box(df_plot, x='mcname', y='weeks', title='雑誌別・連載作品別の合計連載週数')
show_fig(fig)

4.3.3. 作者別の合計連載週数

df_plot = \
    df.value_counts(['mcname', 'creator']).reset_index(name='weeks')
fig = px.box(df_plot, x='mcname', y='weeks', title='雑誌別・作者別の合計連載週数')
show_fig(fig)